Background: Effectively and efficiently diagnosing patients who have COVID-19 with the accurate clinical type of the disease is essential to achieve optimal outcomes for the patients as well as to reduce the risk of overloading the health care system. Currently, severe and nonsevere COVID-19 types are differentiated by only a few features, which do not comprehensively characterize the complicated pathological, physiological, and immunological responses to SARS-CoV-2 infection in the different disease types. In addition, these type-defining features may not be readily testable at the time of diagnosis. Objective: In this study, we aimed to use a machine learning approach to understand COVID-19 more comprehensively, accurately differentiate severe and nonsevere COVID-19 clinical types based on multiple medical features, and provide reliable predictions of the clinical type of the disease. Methods: For this study, we recruited 214 confirmed patients with nonsevere COVID-19 and 148 patients with severe COVID-19. The clinical characteristics (26 features) and laboratory test results (26 features) upon admission were acquired as two input modalities. Exploratory analyses demonstrated that these features differed substantially between two clinical types. Machine learning random forest models based on all the features in each modality as well as on the top 5 features in each modality combined were developed and validated to differentiate COVID-19 clinical types. Results: Using clinical and laboratory results independently as input, the random forest models achieved >90% and >95% predictive accuracy, respectively. The importance scores of the input features were further evaluated, and the top 5 features from each modality were identified (age, hypertension, cardiovascular disease, gender, and diabetes for the clinical features modality, and dimerized plasmin fragment D, high sensitivity troponin I, absolute neutrophil count, interleukin 6, and lactate dehydrogenase for the laboratory testing modality, in descending order). Using these top 10 multimodal features as the only input instead of all 52 features combined, the random forest model was able to achieve 97% predictive accuracy. Conclusions: Our findings shed light on how the human body reacts to SARS-CoV-2 infection as a unit and provide insights on effectively evaluating the disease severity of patients with COVID-19 based on more common medical features when gold standard features are not available. We suggest that clinical information can be used as an initial screening tool for self-evaluation and triage, while laboratory test results should be applied when accuracy is the priority.

Background: Effectively identifying patients with COVID-19 using nonpolymerase chain reaction biomedical data is critical for achieving optimal clinical outcomes. Currently, there is a lack of comprehensive understanding in various biomedical features and appropriate analytical approaches for enabling the early detection and effective diagnosis of patients with COVID-19. Objective: We aimed to combine low-dimensional clinical and lab testing data, as well as high-dimensional computed tomography (CT) imaging data, to accurately differentiate between healthy individuals, patients with COVID-19, and patients with non-COVID viral pneumonia, especially at the early stage of infection. Methods: In this study, we recruited 214 patients with nonsevere COVID-19, 148 patients with severe COVID-19, 198 noninfected healthy participants, and 129 patients with non-COVID viral pneumonia. The participants’ clinical information (ie, 23 features), lab testing results (ie, 10 features), and CT scans upon admission were acquired and used as 3 input feature modalities. To enable the late fusion of multimodal features, we constructed a deep learning model to extract a 10-feature high-level representation of CT scans. We then developed 3 machine learning models (ie, k-nearest neighbor, random forest, and support vector machine models) based on the combined 43 features from all 3 modalities to differentiate between the following 4 classes: nonsevere, severe, healthy, and viral pneumonia. Results: Multimodal features provided substantial performance gain from the use of any single feature modality. All 3 machine learning models had high overall prediction accuracy (95.4%-97.7%) and high class-specific prediction accuracy (90.6%-99.9%). Conclusions: Compared to the existing binary classification benchmarks that are often focused on single-feature modality, this study’s hybrid deep learning-machine learning framework provided a novel and effective breakthrough for clinical applications. Our findings, which come from a relatively large sample size, and analytical workflow will supplement and assist with clinical decision support for current COVID-19 diagnostic methods and other clinical applications with high-dimensional multimodal biomedical features.

Coal Workers’ Pneumoconiosis (CWP) is the primary occupational disease in China. However, information about the definite prevalence of CWP is only partially available. The aims of our study were to assess the prevalence characteristics of CWP in a state-owned coal mine, evaluate the effects of control measures and develop further preventive strategies for CWP. The total study population included 495 cases who were diagnosed with CWP from the construction of this coal mine to the end of October 2014. Individuals’ information, including duration of dust exposure, job titles, age as first diagnosis, stages of CWP, CWP progress, complications with pulmonary tuberculosis, death and others were collected and analyzed. The results showed that 71.11% of 495 CWP cases were stage I and 90.71% were involved in tunneling or coal mining. The mean dust exposure period in CWP patients was 26.7 years, the mean latent period was 29.3 years and the mean diagnosed age was 50.3 years old. The proportion of CWP diagnosed after ending dust exposure were remarkably increased with the time passing. Among the CWP cases, 36 (7.27%) were complicated with pulmonary tuberculosis. The mortality of patients with stage III was the highest (60.71%) (p < 0.0001). Our data obviously show that more strict policies to protect coal miners are needed to be implemented in China, especially for tunneling and mining workers.

To investigate the mortality probability, life expectancy of coal workers’ pneumoconiosis (CWP), and related factors of life expectancy, a total of 495 patients with CWP were diagnosed and reported from 1963 to 2014 in a state-owned mine in the east of China. The life table method, log rank method, and Cox regression model were used for survival analysis. 95 out of 495 CWP died during this period. The mortality rate was 19.19%. The average life span was 12.1 (0.0–33.2) years and average death age was 57.4 (33.0–83.0) years. The life table indicated that overall mortality probability increased with the age of CWP patients. Life expectancy of CWP patients was prolonged to 4.3, 1.4, 1.2, and 1.4 years without death caused by pneumoconiosis, tuberculosis, lung cancer, and pulmonary heart disease respectively. The survival curve of CWP patients without pulmonary tuberculosis was higher (average 37.9 years) than patients with pulmonary tuberculosis (average 34.1 years). There was significant difference observed (χ² = 6.196, p < 0.05). Three risk factors that include initial dust exposure year, age of onset, and first diagnostic stage were put into the Cox regression model for evaluation. The data indicated that prevention and treatment of CWP complication is important to improve patients’ survival rates.

This study aims to describe the characteristics and trends of pneumoconiosis in the Jiangsu Province, China, and provide information for the occupational diseases control. We collected and analyzed the data of pneumoconiosis cases reported annually from 2006 to 2017. The information of the cases mainly includes case distributions, clinical types and stages, enterprise types and scales, as well as diagnosis age and exposure duration. A total of 9243 pneumoconiosis cases were reported between 2006 and 2017, among which silicosis and coal workers’ pneumoconiosis accounted for the vast majority (87.5%). The incidence of pneumoconiosis was relatively higher in Wuxi, Yancheng, Suzhou and Xuzhou, compared to the other district. Most pneumoconiosis cases occurred in the state-owned (58.4%) and collective enterprises (23.8%). Most cases worked in industries related to geology and coal production. The median exposure duration and diagnosis age of the total pneumoconiosis cases was 13.2 and 61.0 years, respectively. Therefore, more measurements are needed to control pneumoconiosis in the Jiangsu Province.

Elevated lead absorptions are hazardous factor in lead-related workers. Previous studies have found its toxic impacts on nervous, circulatory, and metabolic systems. We hypothesized that alteration of miRNAs profile in plasma was closely associated with lead exposure. We analyzed to identify lead-related miRNAs of workers occupationally exposed to lead. Microarray assay was performed to detect plasma miRNA between workers with high and minimal lead exposure in the discovery stage. The following prediction of miRNAs' candidate target genes was carried out by using miRecords, STRING and KEGG databases. We finally identified four miRNAs significantly associated with high level of blood lead. miR-520c-3p (* P = 0.014), miR-211 (* P =0.019), and miR-148a(* P = 0.031)were down-expressed in workers with high lead exposure and with a high blood lead level, while the miR-572(* P = 0.027) displayed an opposite profile. Functional analysis of miRNAs displayed that these miRNAs could trigger different cellular genes and pathways. People under chronic lead exposure had a diverse "fingerprint" plasma miRNA profile. Our study suggested that miR-520c-3p, miR-211,miR-148a and miR-572 were potential biomarkers for lead susceptibility in Chinese.